Steady State of microemulsions in shear flow
Extracted source text ยท arXiv cond-mat/0004282
\documentstyle[twocolumn,prl,aps,epsfig]{revtex}
%\documentstyle[preprint,prb,aps,epsf]{revtex}
\def\inseps#1#2{\def\epsfsize##1##2{#2##1} \centerline{\epsfbox{#1}}}
\begin{document}
\draft
\input{psfig}
%\twocolumn[\hsize\textwidth\columnwidth\hsize\csname @twocolumnfalse\endcsname]
\title{Steady State of microemulsions in shear flow}
\author{F. Corberi}
\address{Istituto Nazionale per la Fisica della Materia,
Unit\`a di Salerno {\rm and} Dipartimento di Fisica, Universit\`a di Salerno,
84081 Baronissi (Salerno), Italy }
\author{G. Gonnella and D. Suppa}
\address{Istituto Nazionale per la Fisica della Materia, Unit\`a di Bari
{\rm and} Dipartimento di Fisica, Universit\`a di Bari, {\rm and}\\
Istituto Nazionale di Fisica Nucleare, Sezione di Bari, via Amendola
173, 70126 Bari, Italy.}
\date{\today}
\maketitle
\begin{abstract}
Steady-state properties of microemulsions in shear flow
are studied in the context of a Ginzburg-Landau free-energy
approach. Explicit expressions are given for
the structure factor and the time correlation function at the
one loop level of approximation.
Our results predict
a four-peak pattern for the structure factor, implying the simultaneous
presence of interfaces aligned with two different orientations.
Due to the peculiar interface structure
a non-monotonous relaxation of the time correlator is also found.
\end{abstract}
\pacs{PACS numbers: 61.20.Gy; 82.70.-y; 83.50.Ax}
Self-assembling amphiphilic systems are binary or ternary
mixtures with a surfactant forming interfaces between the other
fluid components.
These systems show a very
rich phase behavior
with ordered and structured disordered
phases at different temperatures and relative concentrations
\cite{GR}.
Particularly interesting for the applications is
the microemulsion phase where
coherent domains of oil and water
on scales between 100 and 1000 $\AA$ form
disordered isotropic intertwined structures on larger scales \cite{DeG}.
While a satisfactory comprehension of the equilibrium behavior of amphiphilic
systems is now available \cite{GR}
the dynamical features are far less understood.
In particular, in microemulsions, the existence of correlated mesoscopic
structures makes the effects of imposed flows very different
than in simple fluids.
Interfaces between oil and water are
elongated in the flow direction and an excess stress appears in the system.
This can give
a peculiar dynamics with unusual relaxation properties
and a non-Newtonian rheological behavior \cite{larson}.
Moreover, from a statistical
mechanics point of view, complex fluids in external flow are
an interesting example of an
out of equilibrium
driven system \cite{zia}.
In this Letter we study the behavior of microemulsions in shear flow
with a continuum free-energy approach similar
to that used by Fredrickson and Larson \cite{Fre}
for the study of rheology
of block copolymers and by Onuki and Kawasaki \cite{onu} and others
\cite{miln} for the critical properties.
A Ginzburg-Landau model
for describing
the rheological behavior of ternary mixtures was
considered
by P\"atzold and Dawson
\cite{Daw1}.
However the results of \cite{Daw1} are based on an
evaluation of the structure factor
obtained numerically or analytically only in the limit
of vanishing shear rate $\gamma$.
Here we solve exactly the model for any value of $\gamma $ showing
that in an intermediate shear range
the structure factor is characterized by
four pronounced peaks in the plane
of the shear and flow directions. This
new phenomenon can be interpreted in terms of
a complex spatial pattern where interfaces with two different
orientations coexist;
their relative abundance is tuned
by $\gamma$.
We also consider the behavior
of the two-time correlator.
Due to the coupling between the flow and the interface structure,
this function has a remarkable non monotonic behavior
characterized by the presence of
maxima preceding a fast
asymptotic decay.
The Ginzburg-Landau free-energy generally used
to describe the equilibrium properties of ternary mixtures is \cite{GS}
\begin{eqnarray}
{\cal F}\{\varphi\} = \int d^3 x
\{\frac{a}{2} \varphi^2 + \frac{b}{4!} \varphi^4
&+& \frac{1}{2}(g_0 + g_2 \varphi^2) \mid \nabla \varphi \mid^2
\nonumber\\
&+& \frac{c}{2} ( \nabla^2 )^2 \varphi \}
\label{eqn1}
\end{eqnarray}
where the scalar field $\varphi$ represents the concentration
difference between oil and water components.
The amount of surfactant present in the system is related to
the value of $g_0$ which can be negative,
favoring the appearing of interfaces.
The term proportional to $c$ makes stable the free energy at large
momenta and weights the curvature of interfaces.
The quadratic part of the free-energy
in the disordered phase with $a>0$ and $g_0^2 < 4 a c$ gives a suitable
description of the microemulsion phase;
in particular the space correlation function
\begin{equation}
G(r) \sim {\frac{e^{-r/\xi}}{r}} \sin{ \frac{2\pi r}{d}}
\label{eqn2}
\end{equation}
well fits the experimental data \cite{TS,W,GS2,others}.
The characteristic length
$ d=2 \pi (1/2 \sqrt{a/ c} - g_0 /4 c)^{-1/2}$
represents the size of coherent oil or water regions
which are correlated up to a distance
$\xi= (1/2 \sqrt{a/ c} + g_0 /4 c)^{-1/2}$.
At negative $g_0$,
the structure factor obtained by Fourier transforming Eq.~(2)
has a peak at a finite value of the momentum given by $k_M=\sqrt{|g_0|/2c}$.
Finally, the non-linear terms in eq.(1)
describe the possible effects due to mode coupling and become
more important in approaching the phase boundary.
The kinetic behavior of the mixture, neglecting hydrodynamical
effects \cite{nota1}, is described
by the convection-diffusion equation \cite{CH}
\begin{equation}
\frac {\partial \varphi} {\partial t} + \vec \nabla (\varphi \vec v) =
\Gamma \nabla^2 \frac {\delta {\cal F}}{\delta \varphi} + \eta
\label{eqn2}
\end{equation}
where the velocity field is given by
\begin{equation}
\vec v = \gamma y \vec e_x
\label{eqn4}
\end{equation}
$\gamma$ being the shear rate
and $\vec e_x$ the unit vector in the flow direction.
The stochastic term
$\eta$ is a gaussian white noise, representing thermal fluctuations,
with zero mean and correlation
$\langle \eta(\vec r, t) \eta(\vec r', t')\rangle
= -2 T \Gamma \nabla^2 \delta(\vec r - \vec r') \delta(t-t')$,
as required by the fluctuation-dissipation theorem,
where $\langle ...\rangle $ means the ensemble average. Here
$\Gamma$ is a mobility coefficient and $T$ the
temperature of the heat bath in contact with the fluid.
The cubic terms appearing in
the functional derivative of Eq. (3) will be treated in the one-loop
approximation which describes appropriately disordered phases.
The resulting equation for the Fourier components
$\varphi(\vec k, t)$ is given by
\begin{eqnarray}
{\label{eqn5}}
\frac {\partial \varphi(\vec k,t)} {\partial t}&-&
\gamma k_x \frac {\partial \varphi(\vec k,t)} {\partial k_y}
= - \Gamma k^2 \left [ X(k)+g_2 k^2 S_0(t) \nonumber \right . \\
&+& \left . \left( \frac{b}{2}S_0(t)+ g_2 S_2(t) \right) \right ]
\varphi(\vec k,t) + \eta(\vec k, t) \hspace{.01cm}
\end{eqnarray}
where $X(k)= c k^4 + g_0 k^2 +a$,
$S_0(t)$ and $S_2(t)$ are defined self-consistently
by the relations
\begin{equation}
S_0(t) = \int _{|\vec k|<\Lambda} \frac {d\vec k}{(2\pi)^D} C(\vec k,t)
\end{equation}\begin{equation}
S_2(t) = \int _{|\vec k|<\Lambda} \frac {d\vec k}{(2\pi)^D} k^2 C(\vec k,t)
\label{eqn6}
\end{equation}
and $\Lambda$ is a phenomenological ultraviolet cut-off.
Starting from Eq.(5)
a straightforward calculation \cite {Daw1,corb} give
the dynamical equations for the various correlation functions \cite{corr}.
We first focus on the properties of the structure factor
$\langle \varphi(\vec k, t)\varphi(-\vec k, t)\rangle $.
Using the method of characteristics
the $t \rightarrow \infty$ stationary expression
of the structure factor $C(\vec k)$ can be calculated as
\begin{equation}
C(\vec k) = \int_0^\infty dz 2 \Gamma T \vec K^2(z)
e^{-\int_0^z ds 2\Gamma
{\vec K^2(s)}[a^R+g^R \vec K^2(s) +c \vec K^4(s)]}
\label{eqn6}
\end{equation}
where $\vec{K}(s)= (k_x,k_y+ \gamma k_x s,k_z)$, $ \vec K^4
\equiv (\vec K^2)^2$,
$a^R=a+b/2S_0^\infty + g_2 S_2^{\infty}$, $g^R= g_0 + g_2 S_0^{\infty}$, and
$S_0^{\infty}$ and $S_2^{\infty}$ are the limit for $t\to \infty $ of
$S_0(t)$ and $S_2(t)$.
\begin{figure}
\epsfig{file=fig1.ps,bbllx=70pt,bblly=45pt,bburx=553pt,bbury=800pt,
width=9.8 cm,height=16 cm,clip=}
\caption{The projections of the structure factor on the planes
$k_z = 0 $ (left column) and $k_y = 0 $ (right column) at shear rates
$\gamma = 0.25, 2, 4, 8$ (from top to bottom).
The axis $k_x$ points towards the right. The vertical scale is the same
in all the pictures.}
\end{figure}
Since $\xi$ and $d$ are the physical quantities to be compared with
experimental data, the model parameters are tuned \cite{Daw1}
with the following procedure: We first calculate the one-loop
renormalization of the bare coefficients $a$ and $g$ without shear
and express the correlation lengths
$\xi$ and $d$ in terms of the renormalized parameters.
Then, for a given choice of $\xi$ and $d$,
we invert the above mentioned relations and compute
the values of $a,g_0$ to be used in the case with shear.
In the following we show results for the case $\xi=7, d=5$.
The other parameters are set to the values
$c=2$, $b = g_2 = T = 1$.
Similar results have been obtained for different choices of parameters
with $d / \xi$ in the range between 0.5 and 2.5.
The effects of the flow on $C(\vec k)$ are
shown in Fig.~1, where the
projections of the three-dimensional structure factor on the planes
$k_y=0$ and $k_z=0$ are plotted at different $\gamma$.
At $\gamma =0$ the structure factor is isotropic and its shape
on each cartesian
plane is that of a circular volcano with radius $k_M$.
When shear is applied, the projection on the $k_x=0$ plane
gives the same results of the $\gamma =0$ case because
the velocity field is along the $x$ direction.
On the $k_z=0$ plane, instead, this pattern is
progressively distorted as the shear rate is increased.
For small $\gamma $ (see the case $\gamma =0.25$ in the left column of
Fig.~1) the edge of the volcano becomes elliptical and slightly
depressed along the $k_x=-k_y$
direction so that two weak symmetric maxima
located at $\theta= \arctan (k_y/k_x)=\pi/4, 3\pi/4$ are developed.
This is confirmed by a small $\gamma $ expansion of Eq.~(\ref{eqn6}).
As $\gamma $ is increased each of these two maxima
is splitted into
two peaks, as shown in Fig.~1 at $\gamma=2$ and $\gamma =4$.
By increasing the shear rate, the maxima located at
$k_x\simeq 0$ become comparatively more important while the other peaks
rotate clockwise and decrease their amplitude linearly in $\gamma $
until they disappear. Indeed, in
the limit $\gamma \to \infty$,
since terms proportional to powers of
$\gamma k_x$ damp the exponential term on the r.h.s. of Eq.~(\ref{eqn6}),
only the maxima of $C(\vec k)$ with $k_x=0$ survive.
Fig.~1 shows that this is already observable
for $\gamma =8$. Then, by letting
$k_x=0$ in Eq.~(\ref{eqn6})
we find the peaks located at $k_y = \pm k_M$.
The description is completed with
the projection of the structure factors on the plane
$ k_y = 0$ shown in the right column of Fig.~1.
Here two peaks at $k_x = 0, k_z = \pm k_M $,
are observed to become sharper and sharper as the
shear is increased.
A peak of $C(k)$ is generally interpreted as the signature of
a characteristic length in the system which is proportional to the inverse
of its position. In this case, since the system is not isotropic,
to each maximum one associates three lengths, one for each space direction.
Due to the symmetry $\vec k\to -\vec k$ only the peaks
not related by reflection around the origin can be considered.
For large shear the existence of a single couple of maxima
at $k_x=0$ signals that interfaces are aligned along the flow.
In the transverse directions the characteristic lengths are the same as
without shear.
For intermediate values of $\gamma $
the additional peaks at $(\tilde k_x, \tilde k_y, \tilde k_z)$
reveal the
presence of interfaces oriented with an angle
$\alpha = \arctan (-\tilde k_x/ \tilde k_y)$ with respect to the flow, besides
those aligned along the $x$ direction.
As $\gamma $ is
increased these features are progressively tilted in the direction
of the flow while their relative abundance diminishes,
as suggested by the behavior of the maxima with $k_x\ne 0$
previously discussed.
The existence of a four-fold peaked
structure factor has been also reported in scattering experiments on
segregating mixtures \cite{quattropicchi,lcg}.
For small $\gamma $ it is interesting to note that, although the
microemulsion is almost isotropic, the depression of
$C(\vec k)$ along $k_x=-k_y$ indicates a slight predominance of
interfaces directed against the flow at $\alpha \simeq 3\pi/4$.
\begin{figure}
\epsfig{file=fig2.ps,bbllx=30pt,bblly=40pt,bburx=530pt,bbury=500pt,
width=8 cm,height=6 cm,clip=}
\caption{The stress tensor as a function of the shear rate $\gamma$.}
\end{figure}
Stretching of domains requires work against surface tension and results
in an increase $\Delta \eta$ of the viscosity \cite{Onu}.
The excess viscosity is generally defined
as $\Delta \eta = \sigma_{xy}
/ \gamma$ with the shear stress
given by
$\sigma_{xy} = -\int
\frac {d\vec k}{(2\pi)^D} k_x k_y (g^R + 2 c k^2) C(\vec k) $
\cite{Daw1}.
%Another experimentally relevant
% quantity is the first normal
%stress $N_1 = \int
%\frac {d\vec k}{(2\pi)^D} (k_y^2- k_x^2)(g^R + 2 c k^2) C(\vec k) $.
In fig.~2 $\sigma_{xy}$ is plotted as a
function of $\gamma$. Shear thinning is observed.
For small $\gamma $, $\sigma_{xy}$ grows linearly with
$\gamma $
as also found in \cite{Daw1}.
This behaviour is observed in correspondence of a structure factor
similar to that shown in Fig.~1 at $\gamma =0.25 $.
When $C(\vec k)$ develops four maxima at $k_z=0$,
$\sigma_{xy}$ keeps increasing with a much smaller
effective exponent consistent with the value $0.13$.
For very large $\gamma$, when the peaks at
$k_x=0$ alone survive, $\sigma_{xy}$ decreases to zero
because the excess stress vanishes when the interfaces are
aligned with the flow. A similar behavior is shown by
the first normal
stress $N_1 = \int
\frac {d\vec k}{(2\pi)^D} (k_y^2- k_x^2)(g^R + 2 c k^2) C(\vec k)$
\cite{notan1}.
Next we consider the dynamical properties of the microemulsion phase.
In the steady state the two-time correlation function
${\cal D}(\vec k,\vec k',t) =
\langle \varphi(\vec k, t)\varphi(\vec k', 0)\rangle=
D(\vec k,t)\delta \left( \vec K(t)+k' \right )$ \cite{notadelta}
can be computed through
\begin{equation}
{\label{eqn16}}
D(\vec k,t) =C\left(\vec K(t)\right)
e^{- \int_0^t \Gamma
{\vec K^2(s)}[a^R+g^R \vec K^2(s) +c \vec K^4(s)] ds}
\label{timecor}
\end{equation}
$D$ is the product of two terms: the structure factor
evaluated at the translated momentum $\vec K(t)$
and an exponential term.
For a given
$\vec k = \vec K(0)$, $\vec K(t)$ moves along the $k_y$ direction upward
or downward depending on the sign of $k_x$, as represented
schematically in the lower inset
of Fig.~3.
Therefore, when $\vec K(t)$ crosses the edge of the structure factor,
$C\left(\vec K(t)\right)$ shows a pronounced maximum.
This feature is peculiar to microemulsions
where the presence of interfaces gives the patterns
of structure factor discussed above
and cannot be observed in simple fluids.
The exponential term in Eq.~(\ref{timecor}) contains a
polynomial of $t$ and behaves as
$\exp (-\Gamma \gamma^6 k_x^6 c t^7/7)$ for long times.
However, due to the presence of negative coefficients,
the decay can be much slower at earlier times,
as it can be observed in the upper inset of Fig.~3.
In particular one can show that an inflection point is developed
at $\tan \theta = k_x/k_y=-\gamma t$ in the sectors with $k_x k_y<0$.
Other inflection points come out, even in the other sectors,
due to the negative value of $g^R$. Such a slow decay can preserve
the observation of the maximum of $C\left(\vec K(t)\right)$ in
the full time correlator (\ref{timecor}).
In Fig.~3 the behaviour of $D$ is shown for three typical values
of $\vec k$. In case A, $\vec K(t)$ does
not intersect the volcano of the structure factor so that $D$ decays
monotonously. When the edge of $C(\vec k)$
is crossed once or twice, as in cases B and C respectively,
$D$ is characterized by a corresponding number of peaks.
This rich relaxation behavior
has never been described before to our knowledge and
could be searched for in experiments \cite{notago}.
\begin{figure}
\epsfig{file=fig3.ps,bbllx=60pt,bblly=40pt,bburx=570pt,bbury=440pt,
width=9.0 cm,height=7 cm,clip=}
\caption{The behavior of the time correlation function of Eq.~(9)
is shown for the three values of the wave-vector corresponding
to the positions
A, B, C in the lower inset of the figure. The elliptical shape
in this inset represents schematically the contour plot of the structure
factor in the plane $k_z=0$. Solid, dotted, and dashed lines correspond
respectively to the cases C, B, and A.
In the upper inset the exponential factor
in the r.h.s. of Eq.~(9) is plotted for the same three values ok $\vec k$.}
\end{figure}
In conclusion,
we have presented explicit expressions for the steady-state
structure factor and the
time correlation function of a Ginzburg-Landau model describing
mixtures of oil, water and surfactant in the microemulsion phase
under the action of a shear flow.
The structure factor exhibits a four peak pattern which implies a non-uniform
orientation of interfaces with two preferred directions.
The time correlator, reminiscent of the interface structure,
shows maxima superimposed on a global
decay.
All these predictions can be tested experimentally and show that
together with the equilibrium behavior also the dynamics of complex
fluid is characterized by a very rich phenomenology.
~\\
We thank Kenneth Dawson for helpful discussions.
F.C. is grateful to M.Cirillo and R. Del Sole
for hospitality in the University of Rome.
F.C. acknowledges support by the TMR network contract ERBFMRXCT980183
and by MURST(PRIN 97).
G.G. acknowledges support by MURST (PRIN 97).
\begin{references}
\bibitem{GR}
For a review see, e.g., G. Gompper and M. Schick,
{\it Self-assembling amphiphilic systems}
in {\it Phase
Transitions and Critical Phenomena}, edited by C. Domb and J.L. Lebowitz
(Academic Press, New York, 1994), vol.16;
K.A. Dawson in {\it Structure and Dynamics of Strongly Interacting
Colloids and Supramolecular Aggregates in Solution}, edited by
S.-H.Chen at al. (Kluwer Academic Publ. 1992).
\bibitem{DeG}
P.G. de Gennes and C. Taupin, J. Phys. Chem. {\bf 86}, 2294 (1982).
\bibitem{larson}
See, e.g., R.G. Larson, {\it The structure and Rheology of Complex fluids}
(Oxford University Press, New York, 1999).
\bibitem{zia}
B. Schmittmann and R.K.P. Zia, {\it Statistical Mechanics of Driven
Diffusive Systems}
in {\it Phase
Transitions and Critical Phenomena}, edited by C. Domb and J.L. Lebowitz
(Academic Press, New York, 1995), vol.17.
\bibitem{Fre}
G.H. Fredrickson, J. Chem. Phys. {\bf 85}, 5306 (1986);
G.H. Fredrickson and R.G. Larson {\it ibid.} {\bf 86}, 1553 (1987).
\bibitem{onu}
A. Onuki and K. Kawasaki, Ann. Phys. (N.Y.) {\bf 121}, 456 (1979).
A. Onuki, J. Chem. Phys. {\bf 87}, 3692 (1987).
\bibitem{miln}
M.E. Cates and S.T. Milner, Phys. Rev. Lett. {\bf 62}, 1856 (1989).
\bibitem{Daw1}
G. P\"atzold and K. Dawson, Phys. Rev. E {\bf 52}, 6908 (1995);
G. P\"atzold and K. Dawson, J. Chem. Phys. {\bf 104}, 5932 (1996).
\bibitem{GS}
G. Gompper and M. Schick, Phys. Rev. Lett. {\bf 65}, 1116 (1990).
\bibitem{TS}
M. Teubner and R. Strey, J. Chem. Phys. {\bf 87}, 3195 (1987).
\bibitem{W}
B. Widom, J. Chem. Phys. {\bf 90}, 2437 (1989).
\bibitem{GS2}
G. Gompper and M. Schick, Phys. Rev. Lett. {\bf 62}, 1647 (1989).
\bibitem{others}
A. Cappi, P. Colangelo, G. Gonnella, and Amos Maritan,
Nucl. Phys. B {\bf 370} 659 (1992);
P. Colangelo, G. Gonnella, and Amos Maritan,
Phys. Rev. E {\bf 47} 411 (1993).
\bibitem{nota1}
We are also neglecting the motion of surfactant assuming
that its relaxation is fast compared to that of the order parameter
$\varphi$ (see \cite{Daw1}).
\bibitem{CH}
For a review, see A. Onuki, J. Phys.: Condens. Matter {\bf 9} 6119 (1997).
\bibitem{corb}
U. Marini Bettolo Marconi anf F. Corberi,
Europhys. Lett. {\bf 30} 349 (1995).
\bibitem{corr}
Details will be published elsewhere.
\bibitem{quattropicchi}
K. Migler, C. Liu, and D.J. Pine, Macromol. {\bf 29} 1422 (1996).
\bibitem{lcg}
F. Corberi, G. Gonnella, and A. Lamura, Phys. Rev. Lett. {\bf 83} 4057 (1999).
\bibitem{Onu}
A. Onuki, Phys. Rev. A {\bf 35} 5149 (1987).
\bibitem{notan1}
$N_1$ behaves at small $\gamma$ like $\gamma^2$.
\bibitem{notadelta}
Notice the presence of the delta function
$\delta \left ( \vec X(t)+\vec k' \right )$, as opposed
to the usual $\delta (\vec k+\vec k')$, due to the distortion
induced by the flow \cite{onu}.
\bibitem{notago}
It is interesting to observe that an oscillating
behavior of the time correlation function of microemulsions
has been also predicted in the case without shear
when hydrodynamic effects are considered (G. Gompper and
M. Hennes, Phys. Rev. Lett. {\bf 73} 1114 1994).
\end{references}
\end{document}